The use of non-monotonic contracts in a single period game: An experimental investigation
This paper explores whether laboratory students design non-monotonic contracts when they are theoretically optimal in a simple principal-agent context. The principal constructs a contract for three possible outcomes, LOW, MEDIUM, and HIGH, and the agent observes the contract and responds with either an effort level of 0 or 12. Non-monotonic contracts that motivate high effort are optimal for principals because effort only affects the probability of LOW and MEDIUM outcomes. Our experiment has two treatments: one where principals interact with human agents and another with computer agents. Principals only select non-monotonic contracts with a frequency of 15% and 33% in each treatment, respectively. The results are consistent with concerns about equity and agent rationality.
non-monotonic contracts, principal-agent problems, non-monotonicity, experimental economics, contract theory, moral hazard
Flannery, Timothy, and Stephen Roberts. "The use of non-monotonic contracts in a single period game: An experimental investigation." Journal of behavioral and experimental economics 77 (2018): 177-185.